Charset for newsgroups file

Russ Allbery rra at
Wed Nov 3 17:40:09 UTC 2004

Felix E Klee <felix.klee at> writes:

> A bad idea, IMHO: Currently most news readers seems to assume Latin-1
> encoding and people will continue using current readers for years to
> come.  A simple solution to the problem would be mandating that the
> encoding is specified on a line in the newsgroups file (e.g. "\encoding:
> UTF-8").

This would break existing software even more hideously than using UTF-8

> Another solution would be introducing an NNTP command that returns the
> encoding.

We could do this fairly easily, but it really doesn't deal with all of the
underlying issues.  There are also problems with how to handle control
messages, there's inconsistency with the direction that NNTP is going,
there are questions of how to handle non-ASCII group names (rather than
just descriptions), and so forth.

There aren't any good solutions to this, only painful ones, but moving to
UTF-8 as the standard character set is what every other IETF protocol is
doing other than the ones that are staying pure 7-bit ASCII.

Russ Allbery (rra at             <>

    Please send questions to the list rather than mailing me directly.
     <> explains why.

More information about the inn-workers mailing list